Imagine a smart camera that watches a laser weld happening and instantly adjusts the laser if something goes wrong — like a self-driving car but for welding. The system takes multispectral images (seeing both visible light and heat), measures the exact temperature of the weld pool, and uses machine learning to spot defects before they happen. It then automatically corrects the laser power, speed, and gas flow on the fly. The team built a working prototype and tested it on real welding and metal 3D printing lines at factory sites.
Laser welding and laser metal deposition are powerful manufacturing processes, but quality control remains a major headache. Defects like porosity, cracking, or dimensional errors are often caught only in post-production inspection — after the damage is done. Existing monitoring systems typically measure only one variable at a time and cannot automatically correct the process in real time, leading to scrap, rework, and downtime.
The team built a compact multispectral imaging system (visible to mid-wave infrared) combined with an embedded real-time controller and a machine learning-based cognitive control system. The TRL 7 prototype was demonstrated in operational environments for laser welding and laser metal deposition, capable of simultaneously measuring temperature distribution, 3D seam profile, 2D melt pool geometry, surface texture, and process speed — then automatically adjusting laser power, speed, powder flow, gas flow, and spot size.
If you are an automotive manufacturer dealing with weld defects and costly post-production inspections — this project developed a compact multispectral camera and embedded control system that monitors laser welds in real time and auto-corrects process parameters. The TRL 7 prototype was demonstrated in operational welding environments with 12 consortium partners across 7 countries.
If you are an aerospace MRO shop struggling with inconsistent quality in laser metal deposition for turbine blade repair — this project built a closed-loop control system that measures temperature distribution, 3D seam profile, and melt pool geometry simultaneously. It acts on laser power, process speed, powder flow, and spot size to keep every layer within spec.
If you are a metal additive manufacturing company losing parts to porosity and layer defects — this project created a machine learning-based cognitive control system that diagnoses process quality and self-adjusts in real time. The system was validated for laser metal deposition in operational scenarios at end-user facilities.
The project does not disclose per-unit pricing. The total EU research investment was EUR 3,673,157 across 12 partners over 3 years. The modular design means it can be customized for different laser processing setups, which suggests integration costs would depend on your specific application.
The system was designed for real-time closed-loop control with embedded processing, specifically built for high-speed multiple-input/multiple-output control. The prototype reached TRL 7, meaning it was demonstrated in operational welding and laser metal deposition environments, not just lab conditions.
The project involved 12 partners including 5 SMEs and 4 industrial companies. IP arrangements would need to be discussed with the coordinator, ASOCIACION DE INVESTIGACION METALURGICA DEL NOROESTE in Spain. The consortium addressed standardisation issues for the processes and control system.
MAShES was designed under a modular approach, explicitly meant to be easily customizable for different laser processing applications. The system was validated on both laser welding and laser metal deposition, suggesting adaptability to various laser setups. Interoperability with cyber-physical factory systems was also addressed.
The prototype reached TRL 7, demonstrated in operational environments for both welding and laser metal deposition. This is beyond lab testing but may still require engineering work for full commercial deployment in your specific production setup.
The project specifically addressed standardisation issues regarding both the laser processes and the control system itself. It was also designed for compliance with cyber-physical operation in networked factory environments. Based on available project data, specific certifications achieved are not detailed.
The MAShES consortium brings together 12 partners from 7 countries (Germany, Greece, Spain, France, Italy, Sweden, Slovenia), with a healthy mix of 4 industrial companies, 5 research organizations, 1 university, and 2 others. Five partners are SMEs, giving the project a 33% industry ratio. The coordinator is ASOCIACION DE INVESTIGACION METALURGICA DEL NOROESTE, a Spanish metallurgical research association. This broad European spread and strong industrial participation — combined with the fact that the prototype was validated at end-user facilities — suggests the technology has been tested against real manufacturing requirements, not just academic benchmarks.
Contact the coordinator ASOCIACION DE INVESTIGACION METALURGICA DEL NOROESTE (Spain) through the CORDIS project page or project website for licensing and collaboration inquiries.
Want a tailored briefing on how MAShES laser control technology could fit your production line? SciTransfer can connect you directly with the research team and help evaluate the business case for your specific application.
